Ray

Ray is a unified framework designed to scale AI and Python applications. It consists of AI libraries with a basic distributed runtime and can be run from a single machine to a cluster. It contains a variety of components for data, education, hyperparameter setting, reinforcement learning, and model presentation.

RA

Ray

95

ray-project · MLOps & AI Infrastructure

STARS44.2K
FORKS8.2K
COMMITS—
CONTRIBUTORS1.6K
VERSIONray-2.9.3
UPDATEDtoday
PythonDistributedMl InfrastructureApache-2.0

Common commercial products in this category: Weights & Biases, Amazon SageMaker, Databricks

What can it do?
  • You run the Python code across the cluster with the actors (actor) holding status tasks (task) and status information.
  • You perform distributed machine learning training and hyperparameter setting with Ray Train and Tune libraries.
  • You make scalable and programmable model presentation with the Ray Serve component.
Heads up
  • • High number of open issues (3,699)
User reviews
  • ↑ PositiveHacker News

    R.I.P. Ray Bradbury, Author of Fahrenheit 451 and The Martian Chronicles

    “Great author, I am a fan of his Martian Chronicles the most, there are so many short stories of his that I have come across that picking favorites is not fair. Today CATO released an interesting story about Ray and the form of censoring he faced which is one many here would…”

  • • NeutralHacker News

    R.I.P. Ray Bradbury, Author of Fahrenheit 451 and The Martian Chronicles

    “"My favorite Ray Bradbury memory is the time he came to my school to talk about his books and ended up yelling at my English teach for trying to find hidden meaning that wasn't there just for the sake of busywork." Also authors have a subconscious.”

Verified against README · Oct 26, 2026 · source

My note

Alternative pages

More in: MLOps & AI Infrastructure